Quant. MAC Working Group Meeting Book (March 13, 2019)
February 2019 Conference Call
Long Term Objective: • Develop or adopt acceptance criteria for evaluating an alternative method and validate an accompanying workbook • Move towards complete harmonization with ISO statistics • Move towards a modified approach to the ISO statistics • Modify workbook to allow for evaluation of 5 replicates • Combine 2 tier approach • Develop own consensus based acceptance criteria with assistance of the ISPAM statistics committee Proposed Technical Bulletin • Acceptable bias (difference of means) between two methods at a single contamination level falls between a 95% confidence interval of (-0.5, 0.5) • The following table provides guidance on evaluating statistically different bias with individual matrices:
Acceptable number of levels with statistically different bias/matrix
Number of Contamination Levels Evaluated/matrix
3 4 5 6
1 2 2 3
Made with FlippingBook - professional solution for displaying marketing and sales documents online